Computer Science Department and the Electrical and Computer Engineering Department Colloquium Sponsored by ISTeC
The Limitations of Machine Learning in Adversarial Settings

Speaker: Patrick McDaniel, Distinguished Professor, Electrical Engineering & Computer Science, Penn State University

When: 2:00PM ~ 3:00PM, October 16, 2017

Where: Computer Science Building 130

Contact: Indrakshi Ray (Indrakshi.Ray@colostate.edu)

Abstract: Advances in machine learning have enabled to new applications and services to computationally process inputs in previously unthinkably complex environments. Autonomous cars, automated analytics, adaptive communication systems and self-aware software systems are now revolutionizing markets and blurring the lines between computer systems and real intelligence. In this talk, I consider whether the current use of machine learning in security-sensitive contexts is vulnerable to nonobvious and potentially dangerous manipulation. Here, we examine sensitivity in any application whose misuse might lead to harm—for instance, forcing adaptive network in an unstable state, crashing an autonomous vehicle or bypassing anadult content lter. I explore the use of machine learning in this area particularly in light of recent discoveries in the creation of adversarial samples, and posit on future attacks on machine learning. The talk is concluded with a discussion of the unavoidable vulnerabilities of systems built on probabilistic machine learning, and outline areas for defensive research in the future.

Bio:Patrick McDaniel is a Distinguished Professor in the School of Electrical Engineering and Computer Science at Pennsylvania State University, Fellow of the IEEE and ACM, and Director of the Institute for Networking and Security Research. Professor McDaniel is also the program manager and lead scientist for the Army Research Laboratory’s Cyber-Security Collaborative Research Alliance. Patrick’s research focuses on a wide range of topics in computer and network security and technical public policy. Prior to joining Penn State in 2004, he was a senior research staff member at AT&T Labs-Research.